Spoofing attack is one of the main security threats faced by Global Satellite Navigation System (GNSS). If the spoofing signal is regarded as the fault satellite signal, the conventional Receiver Autonomous Integrity Monitoring (RAIM) algorithm can be used for spoofing detection, but its performance is limited by the number of spoofing signals. In this paper, a random traversal RAIM method is proposed for anti-spoofing applications. In the method, the conventional RAIM algorithm is used as the core, and spoofing signals are detected and eliminated by random signal removal and iteration. When there are more than four real signals, the method can effectively eliminate spoofing signals and get correct positioning results by using real signals. So it solves the problem effectively that conventional RAIM algorithm can only detect but not eliminate spoofing signals to get correct positioning results. Finally, simulation and actual signal test results verify the effectiveness of the proposed method.